Calibrating redshift distributions at z>2 with Lyman-α forest cross-correlations

Abstract

We explore the feasibility of using Lyman-$\alpha$ (Ly$\alpha$) forests to calibrate the ensemble redshift distribution of the high-redshift tail ($2 < z < 3$) of photometric galaxies. We use \texttt{CoLoRe} simulations to create mock DESI five-year Ly$\alpha$ forests and Rubin Observatory LSST ten-year photometric galaxy samples up to $z = 3$, and measure the galaxy redshift distribution via their angular cross-correlations. Due to large redshift-space distortions in the Ly$\alpha$ forest, the conventional $n(z)$ estimator for clustering redshifts does not apply, and we develop a theoretical framework to model the angular cross-correlation directly. Using the simulations, we explore the effects of instrumental noise, continuum fitting, and contamination in the Ly$\alpha$ forest, as well as the impact of angular cross-correlation scales ($\theta$) and redshift bin size ($\Delta z$) on the signal-to-noise ratio (SNR) of the measurements. We find that continuum-fitting methods strongly impact the SNR. With our baseline continuum-fitting method, \texttt{LyCAN}, at angular scales $\theta \simeq 10,\mathrm{arcmin}$ and $\Delta z = 0.1$, we measure the cross-correlation signal at $24\sigma$ significance. If the shape of the redshift distribution and the galaxy-bias evolution are well known for $z < 2$, the cross-correlation can constrain the mean redshift of the galaxy sample to $\sigma_z / (1 + \bar{z}) = 0.006$ at a mean redshift of $\bar{z} = 2$. This demonstrates that Ly$\alpha$ cross-correlation is a reliable and promising method to calibrate the high-redshift tails of photometric Stage-IV galaxy surveys.